The Tier-1 Supplier Problem
Why One Anchor Client Is a Fuse, Not a Foundation
In November 2008, General Motors cut production schedules by about 45%. The shock rolled into its tier-1 supplier base within weeks. American Axle & Manufacturing had been pulling around 75% of its revenue from GM for years, and when GM retreated, the company lost its footing inside a single quarter. Dana Incorporated had emerged from a 2006 Chapter 11 only to watch its order book collapse all over again. Both were billion-dollar firms with thousands of engineers on the payroll, and both carried the same flaw: one customer.
Every business runs into a version of this problem sooner or later. Most boards spot it in hindsight. Most diligence decks bury it on slide 47. Peter Drucker wrote in The Practice of Management (1954) that the purpose of a business is to create a customer, and he meant that in the plural, because a single customer is a lease with extra steps and fewer tenant protections.
Once a single account provides more than 30% of revenue, the company is dependent. It doesn’t matter how strategic the relationship looks, how long the contract has been renewing, or how confidently the customer talks about next year’s PO volume. The math stays what it is.
When one buyer crosses that threshold, the company starts working around the customer’s shape instead of its own. Roadmaps tilt toward their feature requests. Sales cycles get planned around their procurement calendar. Smaller opportunities die in the funnel because nobody has the bandwidth to chase them. By the time a customer is 50% of revenue, the supplier is functionally an outsourced department. By 70%, an employee without equity.
A business with one anchor account is brittle, almost beyond redemption, because any shock to the customer passes straight through, undiluted, into the supplier’s P&L.
Feeling set
Clayton Christensen argued in The Innovator’s Dilemma (1997) that successful companies fail because they optimize for their best existing customers. The same dynamic plays out at the operating level: the customer who pays well, who pays on time, who doesn’t haggle over MSAs, and who renews without drama is also the one who will make a leadership team lazy. Pitch decks stop getting updated. Outbound goes cold. The half-finished case study sits in a Notion doc nobody opens. A slower, more comfortable version of the company settles in.
Comfort is the most expensive feeling on a balance sheet. The calendar runs itself, and leadership tells itself there is no time for new business development, which is technically true because the entire org has been structured so they don’t have to. The pipeline calcifies or goes cold. Positioning drifts toward whatever the anchor needs. The network narrows to the four counterparties already on speed dial. The contract becomes a pair of golden handcuffs.
The goal is plain: a customer base wide enough that no single account can break the table. Top customer under 30% of revenue. Top four under 70%. Anything tighter leaves the firm one crisis away from a covenant breach. The suppliers that came through 2008 intact tended to have wider books. Johnson Controls sold into multiple OEMs across multiple regions and absorbed the shock without missing payroll. Delphi, essentially a GM spin-off, filed Chapter 11 and never fully recovered.
Prospecting in reverse
When a company needs revenue, it pitches badly. Discounts creep in. Standards drop. Deals close that shouldn’t, and every term sheet carries the smell of a bad decision.
When a company doesn’t need revenue, the pitch gets sharper on its own, because positioning is clear, pricing holds, sales stops over-explaining, and the customers who sign tend to be better because the team can afford to walk away from the bad ones without thinking about it too hard.
The best time to diversify a customer base is the time nobody feels like diversifying. That’s the whole trap. The work has to happen while the anchor account is still paying on time, because nothing in the quarterly report is forcing it, and someone has to push the org to do it anyway.
There is a real cost to all this. Some operating efficiency goes to switching context across accounts. Margins on smaller customers run thinner. Some mornings the CFO will look at the customer mix and miss the days when everything was simpler.
But: the suppliers that survived 2008 paid that cost for years before they needed it. The ones that didn’t paid a much higher cost in a single quarter. American Axle still exists, but it trades at a fraction of its pre-crash valuation, a smaller and publicly humbled company. Dana Incorporated restructured and moved on. Both spent years untangling the assumption that their biggest customer would always be their biggest customer, and their comfort nearly killed them.
Every operator carries the same assumption humming under the hood. The customer paying half the burn rate this year might cut the contract in December. The MSA that feels permanent is always a QBR away from ending. The moment a company feels set is the moment to start the next pitch. Nobody is immune to the 30% cap: not founders, not Fortune 500 division heads, and not the portfolio companies sitting at the top of a fund’s TVPI.
Selfonomics is designed, built, and backed by Studio Self
We make tech legible.
Reach out: hello@thisisstudioself.com